|
|
Post by jstruggler on Apr 19, 2018 4:59:36 GMT -8
Hello, I've purchased the WeatherBoard and Weather Sensors and successfully setup up all together on my Raspberry Pi 3. I say successfully because I'm getting readings of wind speed, pressure and temperature and ... rain gauge, even when is not raining. My rain sensor keeps reporting false positives and sending to wunderground wrong readings. I've attached an image with the graphs that are being produced. Do you know what might being causing it?  Regards and thanks in advance Jorge Domingos |
|
|
|
Post by SDL on Apr 19, 2018 18:29:33 GMT -8
Jorge,
There are so many things that could be going wrong here. Here are some possibilities:
1) You mast isn't very solid and it is blowing in the wind triggering the Rain Bucket
2) You have a bunch of electronic noise. One thing you can do with that is make the pullup resistor (from the RainBucket line - see the spec, to VDD) stronger (lower resistance - say add 10K ohm resistance
3) Something is wired wrong = it looks like it is continuously triggering. Sounds like #2 to me.
BP
|
|
|
|
Post by jstruggler on Apr 20, 2018 0:41:26 GMT -8
Nice hint the 2) option. Can you please POST the link to the documentation that I should use in order to do it?
Thanks
|
|
|
|
Post by jstruggler on Apr 23, 2018 1:10:43 GMT -8
Hello again,
I've been all the weekend around this subject but still without satisfactory results.
Here what I did:
- add 10Kohm pullup resistor between rain bucket and VCC
- Replace the rainbucket cabling and even solder the wires to reduce all possible contact noise
- Read some articles about Raspberry pi GPIO debouncing and notice that you're already considering boucing compensation parameters on your code of 40ms
Question, I found some recommendations to add 0.1uF capacitor between GPIO pin and Ground, didn't try because I didnt have it. Do you recommend it?
Besides that I'm out of options. Is it possible that this is happening only to me? I cann't find any post regarding this specific problem.
Hope you can help.
Regards
|
|
|
|
Post by SDL on Apr 23, 2018 12:43:58 GMT -8
JStrugggler,
Hmm.
Are you sure it is physically stable?
You might go in and increase the debounce time in the software.
Adding the 0.1uF capacitor is not a bad idea. It will reduce the bouncing.
None of our buckets are showing this. Could it be you have a loose ground connector?
As John Shovic, our CTO says:
You can always trust your Mother, but you can never trust your Ground.
BP
|
|
|
|
Post by jstruggler on Apr 27, 2018 2:53:21 GMT -8
Hello,
My mind is blowing with this subject. Still having false positives on rain measurements.
Now I've tried:
- change raspberry's to version 2 and version 1, all with the same result as raspberry 3. So, not the rasp.
- Disconnected de Bucket and leave the 10Kohm charge, same result. So, not de bucket
- changed the software parameters do force pullup GPIO.setup(rainPin, GPIO.IN, pull_up_down=GPIO.PUD_UP). same result
- Tried de 0.1uF capacitor between rainPin and ground, same result
- Tried several GPIO pins, same result
and now:
- disconnected the rainPin on raspberry that comes from WeatherBoard, no more false readings.
Might be a problem on my WeatherBoard that would send some noise to the raspberry rain Pin? Can I replace it somehow?
Thanks
|
|
|
|
Post by SDL on Apr 30, 2018 13:01:38 GMT -8
JStruggler,
I really doubt it is the WeatherBoard. There isn't much to the circuit on the Rain Pin. This is a noise problem. Did you add an additional pullup to the pin?
Also, you say you disconnected the rain bucket. Where did you disconnect the bucket? Right on the WeatherBoard?
BP
|
|
|
|
Post by jstruggler on May 9, 2018 5:12:10 GMT -8
Hello,
You're right, not the weather board, is something on the bucket, somehow is getting false readings. I've reinforced the structure to make it steady, let's see what happens.
I'll keep in touch
Regards
|
|
|
|
Post by jstruggler on May 28, 2018 12:19:14 GMT -8
Solved, added a 0.1uF capacitor between bucket pin and ground. After some weeks of testing, can consider solved.
Thank you for all support
|
|
|
|
Post by SDL on May 29, 2018 10:16:23 GMT -8
Excellent solution. It was a noise problem! I should have suggested the capacitor fix. Filters out the high frequency noise.
BP
|
|
